What Is Crossix? How Veeva Crossix Measurement Works

People search "what is Crossix" for two reasons. Either a vendor or agency has put a Crossix report in front of them and they want to know how much to trust it, or a planner has been told to "add Crossix" to a media plan and is not sure if that means an audience, a pixel, or a study. Both are fair questions. This article explains the company and the mechanics at the level a brand lead needs, without guessing at details that only Veeva can confirm.

What is Crossix and who owns it?

Crossix Solutions was founded as a health data and analytics company headquartered in New York. Veeva Systems, the life sciences software company known for CRM and content management, announced an agreement to acquire Crossix in September 2019 and completed the deal in November 2019. Since then the business has operated under the Veeva Crossix name.

At the time of the acquisition, Veeva described Crossix as a provider of privacy-safe US patient data and an analytics platform for media and marketing effectiveness. That description still fits. Crossix is best understood as a health data network: it connects health data and non-health data under a privacy framework, then uses that connection for two jobs, building audiences and measuring outcomes.

How Crossix measurement works at a high level

The mechanics below describe how health data network measurement generally works. Crossix's specific methods are proprietary, so treat this as the shape of the process, not a specification.

1. Exposure capture. Media partners pass exposure events, through tags or data integrations, into a privacy-safe environment. Publishers and DSPs that work with Crossix often have an existing integration.
2. De-identified linkage. Exposure records are linked to de-identified health data. Neither the advertiser nor the publisher sees which individual patient had which health event.
3. Comparison group. Exposed people are compared with similar unexposed people. How similar, and on which variables, is the methodology question worth asking.
4. Outcome reporting. The report shows differences in health outcomes between the groups, such as new prescriptions or brand starts, usually with a statistical confidence measure.

For HCP campaigns the logic is similar, but the unit is usually the prescriber, identified by NPI, rather than the consumer. The general process for that version is in how to attribute prescription lift to programmatic HCP media.

Crossix audiences vs. Crossix measurement

These are different products that people often blur together in planning meetings.

Veeva's public product page describes audience segments built with a predictive model that uses demographic attributes and does not include identifiable health information. In plain terms, the segment is a propensity model trained on what a condition audience looks like, then applied to media audiences. That is a different thing from a list of known patients, and it should be planned and measured that way.

One structural point: if you buy a Crossix audience and measure the same campaign with Crossix, the audience provider and the measurement provider are the same company. That is common in this category and not a problem on its own, but your analytics lead should know it and consider an independent check, such as a holdout test, on the largest tactic.

What about "Crossix audience day"?

This phrase appears in search data, but I could not find a standard industry definition for it. It may refer to a Crossix event, a webinar, or internal shorthand for an audience segment or measurement day count. If it appears in a proposal, ask the person who wrote it to define it in writing. Do not assume it is a metric.

How to read a Crossix report without overreading it

A Crossix readout usually arrives as a lift figure and supporting detail. These are the checks I would make before quoting it in a QBR.

• Coverage. What share of the brand's market does the underlying data reflect? A lift estimate describes the covered population, not necessarily everyone exposed.
• Match rate. How many exposures could be linked? Low match rates make results noisier.
• Comparison design. Matched control or randomized holdout? Matched controls can carry selection bias, especially with retargeting.
• Window. How long after exposure are outcomes counted? Long windows inflate attributed counts.
• Attributed vs. incremental. A count of outcomes among exposed people is not the same as outcomes caused by media. The distinction is covered in attributed vs. incremental prescriptions.

An existing case study on this site, decoding Rx attribution with Crossix and IQVIA, walks through why two credible vendors can report different numbers for the same campaign.

Crossix compared with other measurement approaches

Crossix is not the only route to outcome measurement. Claims-based providers, EHR-based networks, and lab data providers all support campaign studies, and many brands use geographic tests or marketing mix modeling for TV and cross-channel planning. Each data asset sees a different part of the patient and prescriber picture. The guide to pharma outcomes measurement platforms lays the categories side by side.

Where Crossix tends to come up first is DTC measurement, because a health data network can link consumer exposure to patient outcomes in a way a pure prescriber-level dataset cannot. For HCP programs, brands often compare Crossix with prescriber-level claims options. Neither choice is automatically right. It depends on coverage for your specific brand and channel mix.

Questions to ask Veeva before you sign

1. Which current products cover our needs, and what are they called in the contract?
2. What share of our brand's prescriptions does your data cover, by pharmacy channel?
3. Which of our media partners have an existing integration, and which need new tagging?
4. How is the comparison group built, and can we see pre-period balance?
5. Can we run a randomized holdout with you, and how would the report differ?
6. What is the data lag between the end of a flight and a stable readout?
7. What level of detail can our team access: totals, segments, or deciles?

Practical takeaway

If a Crossix report is already on your desk, find the line that describes the comparison group and the attribution window before you look at the lift number. Write both next to the headline figure in your summary slide. That one habit stops most of the overreading that happens when a lift percentage travels up the org chart without its method attached.

Editorial note. Analysis and frameworks are the author's own and do not represent Acxiom or any current or former employer, client, or named platform. Examples labeled hypothetical or illustrative are not results from real campaigns. Nothing here is legal, regulatory, or medical advice.